Age Biases in AI-Generated Art
Author Information
Author(s): Laura Allen, Wenqian Xu, Mariko Nishikitani, Vaishnavi Atul Patil, Sunil Hule, Dana Bradley
Primary Institution: University of Maryland Baltimore County
Hypothesis
This study investigates age-related biases in AI-generated art.
Conclusion
The study found that AI-generated images depict older individuals in a more negative light compared to younger individuals.
Supporting Evidence
- Older individuals in AI-generated images were depicted with lower brightness and sharpness compared to younger individuals.
- Older people were more likely to be shown smiling and wearing eyeglasses than younger individuals.
- Qualitative analysis revealed older individuals were often portrayed with themes of nostalgia and vulnerability.
Takeaway
The study looked at how AI shows older people in art, finding that it often shows them in sad or outdated ways, while younger people are shown as modern and energetic.
Methodology
The study used a mixed method approach, collecting images from an AI-art generator and analyzing them quantitatively and qualitatively.
Potential Biases
The study may be limited by the biases inherent in the AI model used for generating art.
Limitations
The study's qualitative analysis was based on a randomly selected sample of 76 images, which may not represent all AI-generated art.
Statistical Information
P-Value
p < 0.001
Statistical Significance
p<0.001
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website